Project Details
Projekt
Structural Health Information Patterns (SHIPs) for analysing the change in condition of bridges in an automated digital twin - [TwinSHIP]
Applicant
Professor Dr.-Ing. Thomas Braml
Subject Area
Structural Engineering, Building Informatics and Construction Operation
Term
since 2025
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 562732371
In our own preliminary work, it was established that changes in the condition of bridge structures can generally be determined with the help of Structural Health Information Patterns, SHIPs for short, i.e. with the help of changes in data patterns. The challenge now is to trace the changes in the data patterns detected in existing structures back to their physical and mechanical causes. Knowledge of the effects and influences that can cause a change in data patterns then allows the selection and arrangement of sensors on bridge structures to be optimised. The aim of the research project is to develop and validate an algorithm that automatically generates data patterns and then assigns these to the specific changes in the characteristics of the bridges and the measuring systems. In addition, the aim is to automatically derive which measurement data should be collected, to what extent and at which positions in order to generate the greatest benefit for feature extraction. The work is based (a) on the large amount of data available from the Nibelungen Bridge Worms and the OpenLAB research bridge and (b) on the AI-supported training data prepared from it. The automated and scalable data acquisition, storage and processing is based on the in-house developed and tested asset administration shell configurator BBox. The expected findings form the basis for optimised inspection and monitoring of bridges in the future. The following sub-goals are defined to achieve the overall project goal: - Extension of the SHIP concept with Convolutional Neural Networks (CNNs) - Development of an algorithm for the automatic development of reference training data sets from sensor data - Validation of the real-time analysis on the demonstrators of the priority programme. Overall, the following four research questions are to be answered: 1) Can state changes of bridge be detected using SHIP's? 2) What conclusions can be drawn about the causes of these condition changes? 3) To what extent can monitoring systems be optimised on the basis of condition changes? 4) How can the standardised and automated real-time application be implemented?
DFG Programme
Priority Programmes
